Sven Lorenz

An Unmanned Helicopter for Autonomous Flights in Urban Terrain

This work summarizes a multi-disciplinary research project, focusing on key enabling techniques towards true autonomous flight of small, low flying VTOL UAVs. Research activities cover the flying testbed, a simulation and testing environment, as well as integrated components for onboard navigation, perception, planning and control. Promising results and feasibility demonstrations in flight tests underline the successful domain specific enhancements of approaches based on aeronautical engineering, computer science and mobile robotics. The current approaches pave the way towards further research in improved flight control performance and more system autonomy when a-priori mission uncertainties increase.

A Decoupled Approach for Trajectory Generation for an Unmanned Rotorcraft

A decoupled approach to trajectory generation based on a cubic spline geometry formulation is introduced. The distinct consideration of boundary conditions yields a continuously differentiable trajectory definition such that path tracking errors are minimized during flight. A curvature-based, dimensionless space-filling curve allows to determine a suitable velocity profile along the path for hover-capable vehicles. Tracking of the trajectory is enabled by a conversion between the spline parameters and the arc length of the spline. In the past years, this approach in combination with a suitable trajectory tracking control has been successfully flight tested with an unmanned helicopter.

Optimization-Based Feedforward Path Following for Model Reference Adaptive Control of an Unmanned Helicopter

This paper addresses an optimization based approach to follow a geometrically defined path by an unmanned helicopter. In particular, this approach extends reference model following concepts. Instead of using vehicle dynamics, the optimization is based on the reference model of the controller. This way, we can calculate the time-wise evolution on the path by means of dynamic optimization. The progression on the geometric path, namely the timing law, is defined as a dynamic system subject to an additional virtual control input. The inputs of the reference model and that of the timing law are the decision variables used in the dynamic optimization. It will be shown that this way, an accurate following of the path is possible although the actually identified flight mechanical model is limited to a linear hover model. Furthermore, the approach allows to take constraints on inputs and states into account. Simulation results as well as preliminary flight tests conducted with the ARTIS testbed are presented for two nonlinear and planar paths underlining that good performance and constrains satisfaction can be achieved.

Control of a VTOL UAV via Online Parameter Estimation

Due to their ability to hover and to fly “low and slow” VTOL UAVs are of extreme strategic importance. The task of control of VTOL UAVs is however complicated as they exhibit highly coupled nonlinear dynamics and inherent instability. These intricacies limit the performance of linear control techniques since a single linearized state space model represents only a limited part of the flight envelope. Moreover, in case of miniature rotorcraft UAVs, the problem is often accentuated due to inaccurate linear modeling, noisy sensor measurements and external disturbances. In this paper, a control architecture is presented which extends the validity of a linear optimal full state feedback law via online parameter identification and parameter dependent adaptive control. An Extended Kalman Filter is used for the combined problem of state and parameter estimation. Based on the estimated parameters the state feedback gain is calculated by solving the Riccatti equation for quadratic optimized control online. Online estimates of trim values are added to the inputs to account for varying trim conditions. The parameter identification algorithm is tested with flight data and validated against an identified linear model. The control architecture is tested in Software In The Loop simulation and in realtime with Hardware in the Loop simulation. The robust performance of the control architecture in presence of noisy data, parameter uncertainties, external disturbances and unknown trim conditions indicate the feasibility of such an approach for the control of miniature VTOL UAVs.

Modular Simulation Framework for Unmanned Aircraft Systems

This paper presents the concept and implementation of a simulation framework, capable of simulating a variety of different kind of unmanned aircraft. The concept is designed for various tests, including software-in-the-loop as well as hardware-in-the-loop simulations. The aim of this framework is to ease the simulation setup for different purposes including algorithm development, software testing and control station design. Different problem abstractions and a desired degree of simulation complexity shall be available. A terminology for this particular problem formulation is proposed. Different layers of configuration are identified and a framework is introduced which allows the automatic simulation setup. A library based approach is presented independent of a development language or tool. Subsequently, an example implementation is introduced within a hybrid Matlab/Simulink and C/C++ environment enabling a fast reuse of simulation modules, like sensors, airframe configurations, and environmental setups. The paper concludes with lessons learned from using this framework during the past years.

Robustness Analysis Related to the Control Design of the SHEFEX-II Hypersonic Canard Control Experiment

This paper presents the simulation based robustness analysis w.r.t. parameters and initial conditions, related to the hypersonic flight experiment called SHEFEX-II. The robustness criteria and the parameter limits are obtained from the nominal mission profile, the vehicles configuration, and the results of a previous flight test in 2005. Necessary background knowledge and the nominal mission profile are presented. Simulation results are introduced to define the nominal and the worst case trajectories. Due to the fast changing air density during the reentry flight, simulations with modified initial conditions are used to determine the desired damping. Furthermore, preliminary flight test results are presented and discussed.

Evaluation of Time-Shifted Feedforward Control for Unmanned Helicopter Path Tracking

A novel approach to improve the path tracking performance based on time shifted feed-forward signals is presented. It uses approximately linear closed-loop dynamics through feedback linearization of the ARTIS unmanned helicopter. The flight controller is currently in use to track nonlinear trajectories represented by piece-wise cubic splines. Feedforward signals shifted in time with varying and constant time-shifts are evaluated to improve the suboptimal tracking behavior. The problem is illustrated for a representative path and the process of determining the time shift is discussed. Having runtime performance advantages, the constant variant of time-shifted feedforward signals is chosen for onboard implementation. To validate the approach, it is implemented onboard our flying rotorcraft test bed and flight test results showing the improved tracking performance are presented.

Optical-Aided Aircraft Navigation using Decoupled Visual SLAM with Range Sensor Augmentation

This paper presents an optical-aided navigation method for automatic flights where satellite navigation might be disturbed. The proposed solution follows common approaches where satellite position updates are replaced with measurements from environment sensors such as a camera, lidar or radar as required. The alternative positioning is determined by a localization and mapping (SLAM) algorithm that handles 2D feature inputs from monocular camera images as well as 3D inputs from camera images that are augmented by range measurements. The method requires neither known landmarks nor a globally flat terrain. Beside the visual SLAM algorithm, the paper describes how to generate 3D feature inputs from lidar and radar sources and how to benefit from both monocular triangulation and 3D features. Regarding state estimation, the approach decouples visual SLAM from the filter updates. This allows software and hardware separation, i.e. visual SLAM computations on powerful hardware while the main filter can be installed on real-time hardware with possible lower capabilities. The localization quality in case of satellite dropouts is tested with data sets from manned and unmanned flights with different sensors while keeping all parameters constant. The tests show the applicability of this method in flat and hilly terrain and with different path lengths from few hundred meters to many kilometers. The relative navigation achieves an accumulation error of 1–6 % of distance traveled depending on the flight scenario. In addition to the flights, the paper discusses flight profile limitations when optical navigation methods are used.

Radar-aided optical navigation for long and large-scale flights over unknown and non-flat terrain

For flight automation tolerable to satellite navigation dropouts, this paper presents a simultaneous localization and mapping method based on radar altimeter measurements and monocular camera images. The novelty within mapping is the combination of radar distance and image triangulation. This approach verifies whether the radar measurement fits to a specific horizontal plane in the map, yielding the sub-set of image features that do most probably correspond with the radar measurement. With this map match of the radar altitude, ambiguities in the radar measurement can be resolved. Since unusable radar measurements are suppressed, this method is suitable for positioning in non-flat terrain, e.g. in mountain areas. For matched data, the method estimates a scale correction factor for the image projection rays in order to remove scale ambiguities of the monocular navigation. Together with mapping, vehicle localization is done which is essentially camera resectioning. Localization can be parameterized with the required number of degrees of freedom depending on the availability of additional position sensors. The incremental positioning is tested in kilometer-scale outdoor flights of a 30 kg unmanned airplane as well as in flights with a Cessna 172R equipped with camera and radar sensors. The tests show the benefits of the proposed method in flat and hilly terrain, and demonstrate reduction of accumulation errors down to 2-6% over the distance flown. Some constraints of the method for the altitude range are existent, however it is highlighted that this method will generally work on typical flight profiles.

Open-loop reference models for nonlinear control with applications to unmanned helicopter flight

A novel way to design a nonlinear feedforward controller is introduced. Contrary to input-output linearization and pseudo control hedging techniques, the new formulation proposed is based on the states of the reference system instead on the states of the plant. By reshaping the structure commonly used in literature the similarities between the resulting scheme and what is made in model following control are utilized to get a pure nonlinear feedforward system which is completely decoupled from system outputs. This permits the application of the nonlinear control law when knowledge of the system states is not available. Due to the strict separation of plant and reference dynamics in the proposed approach a stability analysis of the reference system can be simplified. This new control scheme has been used for the control of an unmanned helicopter demonstrating its applicability. Results of these flight-test are presented in this paper.